Multi-Axis Wind Load Simulator

The MAWLS is a unique large-scale dynamic wind effects simulator that imposes dynamic wind pressure in combination with static in-plane shear or uplift forces. The system is designed to accommodate walls, components, or cladding specimens up to 18 by 24 ft. It is envisioned that this equipment can be used by researchers to evaluate strength and serviceability performance of systems at a scale not previously achieved in any other laboratory. This system supports the study of the interaction between static uplift or in-plane shear and time-varying pressure conditions to a level associated with an intense Saffir-Simpson Hurricane Wind Scale Category 5 hurricane or an EF5 tornado.

Special-Lite, Inc. was the primary sponsor of the system. Henry Upjohn (President), Ken Bowditch (R&D Manager) and Bob Nicholson (Consultant) contributed heavily to the design and construction. Additional support was provided by the FSU Florida Catastrophic Storm Risk Management Center and the University of Florida.

Capabilities and Principle of Operation

The MAWLS was designed to apply spatially uniform, time-varying pressures on the full-scale building component surface up to the level of a Category 5 hurricane. The simulator is capable of handling large vertical spanning cladding and component systems for a wide range of leakage conditions. A reaction frame placed around the opening provide the multi-axial feature in which in-plane shear or uplift can be applied in combination with time-varying pressures. The Simulator will be capable of subjecting a building system to an entire hurricane episode in order to replace or reduce the need for simple diagnostic tools, such as building product approval tests adapted from ASTM procedures. Conceptually the system can be thought of as a means to replicate naturally-occurring wind and pressure in a controlled laboratory environment. For example, if a pressure sensor recorded high-fidelity data on the wall of a commercial building in Homestead, FL during Hurricane Andrew, the Simulator could “replay” this pressure sequence in its entirety. The measured and artificially applied loading would be virtually indistinguishable.

Prior to structural testing specimens can be subjected to treatments such impacts from windborne debris and natural or artificial weathering effects (wind-driven rain, and thermal cycling), and accelerated aging. The design criteria for the large-scale multi-axis dynamic wind load simulator were to enable high-fidelity replication of intense wind loading, i.e., Category 5 hurricanes, and EF-5 tornadoes.

Size 7.6 m × 12.2 m
Peak pressure 22 kPA
Peak Air Flow 2832 m3/min (100,000 CFM)
Frequency Response 3 Hz
Actuators – Vertical 30 tons
Actuators – Horizontal 30 tons
Performance Envelope view [pdf]
High Performance Actuator 3D view [dwg]
EFS Building view [dwg]
Blast Wall Footprint and Details view [dwg]

The MAWLS test bay accommodates up to a 24 ft wide, two-story wall or other building envelope component, and it can generate simulated wind surface pressures up to 22 kPA at a through-specimen airflow rate of 2,800 m3/min. The system is highly responsive and repeatable. It can follow large amplitude fluctuating pressure traces having fast frequency response with minimal errors in signal reproducibility. In addition to its capabilities to develop out-of plane pressures, the MAWLS basic test configuration includes three 30-ton hydraulic actuators that can apply both in-plane uplift and in-plane lateral shear loads to the specimen. With its extremely high power reserve, MAWLS can be used for evaluation of complete building systems without need to alter system boundary conditions or to insert materials (i.e. plastic sheets) within other systems to maintain test pressures. Further, MAWLS can validate building envelope designs by testing them to failure to establish ultimate structural capacities.

MAWLS operates in two modes:

  1. Pressure simulation where Dynamic wind pressure is applied to large wall cladding and component systems, such as a commercial rolling door or a wall on a metal building. The machine can also apply static (steady) and pulsating pressure conditions
  2. Velocity simulation where turbulent airflow conditions over a roof deck are simulated to load discontinuous roof systems (e.g., tile and shingles), which is the Dynamic Flow Simulator (DFS)

In deciding on these parameters, the fidelity of the pressure simulator was evaluated by recreating a pressure time history from a scale model. The pressure signal was calculated using the wind pressure coefficient (p) data from generic wind tunnel model m31 archived in the NIST Aerodynamics Database [ref]. The wind tunnel model is a 1:100 scale model of a 3:12 slope gable-roofed building with a 24 m x 38 m (80 ft. x 125 ft.) rectangular pan and a 9.8 m (32 ft.) eave height. The exposure for this wind tunnel model is open country.

Time histories of positive and negative wind pressure with a frequency of 2 Hz at intensity level seven.

This picture shows extent of the volume change that which this can operate at

Dynamic Velocity Simulation

To determine the velocity simulation mode, the pressure chamber is shut off so that air is pulled through the blower from the exterior intake, then passes through the exterior exhaust, and travels into a high-speed wind tunnel section. The wind tunnel section starts with a setting chamber, which reduces the incoming turbulence and then accelerates the airflow through a contraction duct to the target velocity at the entrance to the test section. The cross-section area at the entrance to the test section is 213 cm wide x 38 cm tall (7 ft. x 1.25 ft.). The bottom part of the test section is removable to accommodate roof samples, and it has a dimension of 243 cm long x 182 cm wide (8 ft. x 6 ft.).

Full-scale wind velocity data collected from hurricane Katrina were resampled at four intensity levels with mean wind velocity = 15, 25, 35 and 45 m/s. Fig. 21 shows velocity simulation at the four levels. During wind velocity simulation, the rotational speed of the engine can be adjusted as needed, e.g., levels 1 and 2 correspond to a rotational speed of 1200 RPM, level 3 corresponds to 1450 RPM, and level 4 corresponds to 1700 RPM. In the figure below, the solid black line represents the input (or “target”) velocity signal, and the dash-dot gray line describes the measured signal. It can be seen that the simulator follow the target signal well.